2022.05.19 Meeting Notes

Agenda

• Individual/group updates
• Coverage
• Referee report
• Intel update

Individual updates

LR

• still fixing/updating sparse bvals
• just added flux correction, technically all the bbvals machinery could be remove (but in a future PR)
• regression tests are now in exact agreement
• flux correction and new boundary pieces works well, serial, mpi, cuda but cuda+mpi does not work correctly
• ways to debug: CUDA_LAUNCH_BLOCKING and pack_size=-1
• Issue on missing flux alloc still open

JM

• mostly downstream dev, looking at post pgen mesh work

BR

• downstream dev on particles
• one open PR on particles and one on parallel_scan (for index sorting)

FG

• working on fixing downstream physics

PG

• clang-format
• sparse restarts

Referee report

• received report on paper from two reviewers with minor comments
• PG will address the comments and share an updated draft included updated numbers

Coverage

• coverage stage times out because sparse problems take too long in debug mode
• would need more detailed update of test infrastructure as currently coverage for a test is on or off (no option to just let run with a reduced nlim or similar)
• for now -> will temporarily disable coverage from CI (can still be called manually)

Intel

• next call tentatively scheduled for 31 May at 11am Mountain
• goal
  • status of SYCL backend in Kokkos
  • present Parthenon and downstream codes in more detail
  • report on (fixed?) compile problems